For the 3rd year in a row, @abnormalai is a top 20 company on the @Forbes annual Cloud 100 list! Thank you to Abnormal AI team + customers + partners! More futuristic products & AI stuff coming soon to help protect our customers from the future of cybercrime.

Despite 100x gains in productivity with AI agents There will be *more* security engineers in the future not less

me getting ready for board meeting after anthropic is doing $1B NARR/week

https://t.co/dRTuGSMfL1

The AI conversation is stuck on "faster, cheaper, more efficient." Future models will hold your entire company in their head and connect dots humans physically can't. The intelligence compounds every year. A billion ants can't write Shakespeare.

I love all my customers equally…but I have a special place in my heart for the F500 CISOs that join @AbnormalAI product team to help us build the next gen of AI powered cyber defense!

I enjoyed my recent discussion with Lev Gonick, CIO at @ASU. Arizona State University operates at a massive scale, with 200,000 students and 35 locations globally. They’ve decided to treat AI as a student success strategy, not a threat to academic integrity. A few takeaways: 1)

The scariest AI failure mode isn't sentience. It's an agent that follows your instructions and still does exactly the wrong thing at machine speed. This week, an AI researcher at Meta told her AI agent "confirm before acting." It speedran deleting her entire inbox. She had to

The ADHD Claude code / vibe code experience is starting off building a personal todo app, and 14 hours later at 4am, you are 13 side quests deep (re)building a custom multi agent message bus

I had a great conversation recently with Matt Posid, former CISO at the @CIA and current CSO at @KPMG_US, a 40 billion-dollar global professional services network and one of the Big Four. As a firm trusted with sensitive client and financial data across industries and

https://t.co/Inmv3lRqh2

This happening right now at a small scale. Tomorrow it’s going to happen at an overwhelming scale that humans can’t compete with. We need good AI to fight the bad AI

At @abnormalai we now have 13% of code changes are autonomously initiated by AI (and rising every week)

When building is more easy, more fun and more rewarding than ever, it also becomes clear who the builders are

Since adopting a magical refactoring engine, all deployments ship. My yesterday? Totally wasted endlessly editing trivial stuff

Code bases should be optimized for AI to read vs. Humans. Stop Doing: Human-oriented documentation, CLI design, code organization, naming practices Start Doing: Intent annotations, larger self contained files, importable functions > CLI scripts, structured returns, test-as-spec

State of the art for software development is changing so fast that it feels like every week I look back at the code/architecture/workflows and I think: damn. I’m an idiot…what have I done?

what simulation are we in where the ceos of the top AI labs have last names that look like Sundar: Pitch ai Dario: Ai model Sam: Alt man

This may sound trite, but switching from Claude opus 4.5 to 4.6 with 1M token context window is a major upgrade. The architectural thing i spent 3 hours debugging was one shotted by 4.6 Now the rest of might will be asking it what else I’m missing